601 
completed. It is therefore impossible to directly trace them back into 
the larval period. The conditions for their appearance seem to be 
found in the readjustments taking place at the time of metamorphosis. 
These form a complex of degenerations, new-formations, fusions and 
shiftings, the last notably in 
the articulation and suspenso- 
rium of the jaw. The develop- 
ment of the three tonsils will 
be considered in the sequence 
of increasing complexity. 
The region in which the pre- 
glottideal tonsil appears is the 
simplest and least altered in 
the process of metamorphosis. 
In the larva the glottis opens Fig. 10. The Preglottideal Fold. , Sala- 
smoothly without bordering mandra atra. The section is just through the 
folds or furrows. ‘This is the gpbele end of ho glottis, is a Tile a 
condition in five larvae, 121/,, on the left side. >< 60. 
19, 25, 38, and 44 mm in length 
each. In a 32 mm specimen which despite smaller size is more ad- 
vanced than the longer larvae and has already entered upon its 
metamorphosis, 
the glottis is bor- 
dered by a depres- 
sion upon one 
side and a tubular 
down-growth 
AOE upon the other.) 
EIER At transforma- 
Fig. 11. The same.folds, some sections farther cephalad. <60. tion a crescent or 


> 
s 
horse-shoeshaped 
furrow is developed about the cephalic end of the glottis slit and 
extending around upon the side of the opening. As metamorphosis 
1) This is not the point of origin of the “postbranchial body”. This lies 
at the same general level (left side only), but farther laterad, morphologically 
behind the 6th visceral (4th branchial) arch, and cephalad of this depression 
and the furrow subsequently described. 
Hitton (1911) has described in Amblystoma Sihotake a furrow in this 
general locality which he is inclined to regard as representing a rudimentary 
